Non-contiguous Allocation Strategies Research Articles

On Allocation Algorithms for Manycore Systems With Network on Chip

Single-chip multicore processors and their network on chip interconnection mechanisms have received extensive interest since the early 2000s. The mesh topology is popular in networks on chip. A common issue in mesh is that it can result in high energy consumption and chip temperatures. It has been recently shown that mapping communicating tasks to neighboring cores can reduce communication delays and the associated power consumption and improve throughput. This paper evaluates the contiguous allocation strategy first fit and non-contiguous allocation strategies that attempt to achieve a degree of contiguity among the cores allocated to a job. One of the non-contiguous strategies is a new strategy, referred to as neighbor allocation strategy, which decomposes the job request so that it can be accommodated by free core submeshes and individual cores that have degree of contiguity. The results show that the relative merits of the policies depend on the job's communication pattern.

Improving system performance in non-contiguous processor allocation for mesh interconnection networks

Contiguous allocation of parallel jobs in multicomputers usually suffers from the degrading effects of fragmentation, because it requires that the allocated processors be contiguous and have the same topology as that of the multicomputer's interconnection network. Fragmentation can be avoided by adopting non-contiguous allocation. However, non-contiguous allocation can increase the distances among processors allocated to a job, which can increase the interference among messages of different jobs and increases message contention and delays. This paper suggests a new non-contiguous processor allocation strategy, referred to as Minimum Interference Paging (MIP), for the 2D mesh network. MIP attempts to reduce the distances among processors allocated using a paging variant that chooses a set of processors with the lowest distance between the first and last allocated processors, where the distance is the number of processors between the first allocated node and last allocated node. Using software simulation, we compared the performance of MIP against that of several well-known non-contiguous allocation strategies: paging (0), Multiple Buddy System (MBS) and Adaptive Non-contiguous Allocation (ANCA). The comparative evaluation was conducted using First-Come-First-Served (FCFS) job scheduling, wormhole routing and six communication patterns. These are the one-to-all, all-to-all, Near Neighbour, Ring, Divide and Conquer Binomial Tree (DQBT), and Random communication patterns. The results show that MIP exhibits superior performance in terms of average turnaround time of jobs.

A New Compacting Non-Contiguous Processor Allocation Algorithm for 2D Mesh Multicomputers

In non-contiguous allocation, a job request can be split into smaller parts that are allocated possibly non-adjacent free sub-meshes rather than always waiting until a single sub-mesh of the requested size and shape is available. Lifting the contiguity condition is expected to reduce processor fragmentation and increase system utilization. However, the distances traversed by messages can be long, and as a result the communication overhead, especially contention, is likely to increase. The extra communication overhead depends on how the allocation request is partitioned and assigned to free sub-meshes. In this paper, a new non-contiguous processor allocation strategy, referred to as Compacting Non-Contiguous Processor Allocation Strategy (CNCPA), is suggested for the 2D mesh multicomputers. In the proposed strategy, a job is compacted into free locations. The selection of the free locations has for goal leaving large free sub-meshes in the system. To evaluate the performance improvement achieved by the proposed strategy and compare it against well-known existing non-contiguous allocation strategies, the authors conducted extensive simulation experiments. The results show that the proposed strategy can improve performance in terms of job turnaround times and system utilization.

The Effect of Real Workloads and Synthetic Workloads on the Performance of Job Scheduling for Non-Contiguous Allocation in 2D Mesh Multicomputers

The performance of non-contiguous allocation has been traditionally carried out by means of simulations based on synthetic workloads, and also it can be significantly affected by the job scheduling strategy used for determining the order in which jobs are selected for execution. To validate the performance of the non-contiguous allocation algorithms, there has been a need to evaluate the algorithm's performance based on a real workload trace. In this paper, the performance of the well-known Greedy Available Busy List (GABL) non-contiguous allocation strategy for 2D mesh-connected multicomputers is revisited considering several important job scheduling strategies based on a real workload trace, and the results are compared to those obtained from using a synthetic workload. The scheduling strategies used are the First-Come-First-Served (FCFS), Out-of-Order (OO), and Window-Based job scheduling strategies. These strategies have been selected because they are common and they have been used in related works (Ababneh & Bani-Mohammad, 2011). Extensive simulation results based on synthetic and real workload models indicate that the Window-Based job scheduling strategy can improve both overall system performance and fairness (i.e., maximum job waiting delays) by adopting a large job scheduling window. Moreover, the relative performance merits of the scheduling strategies when a real workload trace is used are in general compatible with those obtained when a synthetic workload is used.

On the performance of non-contiguous allocation for common communication patterns in 2D mesh-connected multicomputers

The communication pattern used by applications can have a major influence on the performance of non-contiguous processor allocation in multicomputers. In this paper, the performance of well-known non-contiguous allocation strategies for 2D mesh multicomputers is re-visited considering several important communication patterns. These are the Near Neighbour, Ring, Divide and Conquer Binomial Tree (DQBT), Fast Fourier Transform (FFT), and Random communication patterns. The allocation strategies investigated are the Greedy Available Busy List (GABL), Multiple Buddy Strategy (MBS), Adaptive Non-contiguous Allocation (ANCA), and Paging(0). They are compared using detailed flit-level simulations. The results show that GABL is overall superior to the remaining non-contiguous allocation strategies. It produces superior average turnaround times and mean system utilization.

Performance evaluation of noncontiguous allocation algorithms for 2D mesh interconnection networks

Several noncontiguous allocation strategies have been proposed for 2D mesh-connected multicomputers. These allocation strategies differ in their ability to detect free submeshes and in the degree of contiguity that exists among the submeshes they allocate to the same job. The previous Adaptive Noncontiguous Allocation (ANCA) policy was evaluated based on a proposed formula that estimates the job execution time when the job is allocated noncontiguous submeshes. Using this formula, simulation results had shown that ANCA could outperform the preceding Multiple Buddy Strategy (MBS). However, the execution times of jobs under noncontiguous allocation depend on message sizes, the number of messages sent, message contention and distances messages traverse. In this paper, we evaluate ANCA for different communication patterns using an event-driven simulator operating at the flit level, which allows for a more realistic evaluation that takes into account the shape of allocation and contention among messages. Moreover, we compare the performance of ANCA with that of other noncontiguous allocation strategies, MBS, Greedy Available Busy List (GABL), and Paging(0). The results show that ANCA is inferior to the remaining strategies, and that GABL has the best performance results, expressed in terms of the average turnaround time and mean system utilization performance parameters.